Dose-effect curves for amphetamine on key-pecking behavior of pigeons maintained by two-component multiple schedules of shock postponement were determined. During the first experimental phase the response-shock interval (RS) was held constant and the shock-shock interval (SS) varied. Under these conditions, shock rate was greater during the component with the shorter SS. However, response rates were comparable in both components. Also, the magnitude of the response rate increments caused by appropriate doses of amphetamine during both schedule components were similar. During the second experimental phase, the SS was held constant and the RS varied. As a consequence, baseline response rate was considerably lower in the component with the longer RS than in the short RS component. Shock presentations were also less frequent during the former than during the latter component, but the differences in shock rate between the components were comparable or smaller than those observed during the first experimental phase. Under these conditions, the effects of amphetamine in the two schedules components were markedly different, lower response rates being considerably more increased than higher rates.